Emulators for Scarce and Noisy Data: Application to Auxiliary-Field Diffusion Monte Carlo for Neutron Matter

Area of Science:

• Nuclear Physics
• Astrophysics
• Computational Physics

Background:

• Understanding the equation of state (EOS) of neutron star matter is crucial for interpreting astrophysical observations.
• Accurate EOS requires propagating uncertainties from nuclear interactions, but this is computationally intensive.
• Existing methods like Quantum Monte Carlo are precise but too slow for uncertainty propagation.

Purpose of the Study:

• To develop fast emulators for neutron matter EOS calculations.
• To enable direct propagation of uncertainties from nuclear Hamiltonians to the EOS.
• To provide robust uncertainty estimates for astrophysical data analysis.

Main Methods:

• Employing parametric matrix models to create emulators for auxiliary-field diffusion Monte Carlo (AFDMC) calculations.
• Using these emulators to propagate uncertainties in nuclear Hamiltonian coupling constants directly to the EOS.
• Incorporating effective field theory truncation uncertainties into the analysis.

Main Results:

• Developed fast emulators for neutron matter EOS calculations.
• Successfully propagated uncertainties from nuclear interactions to the EOS.
• Achieved robust uncertainty estimates suitable for astrophysical applications.

Conclusions:

• The developed emulators significantly speed up EOS calculations, making uncertainty propagation feasible.
• This approach allows for direct constraints on nuclear interaction coupling constants using astrophysical observations.
• Enables novel applications in astrophysics and nuclear physics research.